|
MAIN PART.
Research objectives are planned:
1. Analysis of the influence of various dominant disturbing factors on the accuracy of the shaping process.
2. Construction of models for the formation of processing errors. Conducting their comparative analysis.
3. Development of new methods of adaptive control of the accuracy of processing a batch of parts under conditions of initial a priori uncertainty, including under conditions of obtaining incomplete information in real time.
4. Development of a system for automated selection of control options and control of the accuracy of processing a batch of parts
5. Development of a software module for mathematical modeling of methods of adaptive pulsating adjustments by correcting variable increments of the level of dimensional adjustment of equipment.
At the stage of technological preparation of production and when organizing software and mathematical support for machine modules, it is effective to use a system for automated selection of options for monitoring and controlling the accuracy of processing a batch of parts, implemented on the basis of experimental research and new methods of controlling processing accuracy developed by the author [2].
Fig. 1. Correction of the error by changing the 3D model.
Under conditions of initial (a priori) uncertainty, it is advisable to use methods of adaptive adjustments of correcting increments, especially when incomplete information about the output parameters of the cutting process is received.
When implementing automatic sub-adjusters, it is necessary to take into account the rate of displacement of the level of the dimensional adjustment of the equipment to regulate the measurement frequency.
The process of synthesizing the model for the formation of the processing error was formalized in the form of an algorithm implemented on a PC as a software module. The procedure for the automated synthesis of a model for the formation of a processing error uses well-known and well-tested techniques and allows you to determine the proportion of deterministic, random with correlated values and actually random components. An analysis of the stages of mathematical modeling of the formation of processing errors is carried out, including the determination of laws and parameters of the distribution of experimental data, the influence of disturbing factors on the processing accuracy is investigated.
Taking into account the results of known studies, it is substantiated that the initial (a priori) uncertainty of the shaping process largely determines the accurate prediction (prediction) of dimensional deviations in subsequent processing cycles in real time for the manufacture of a batch of parts with adjustments. High requirements for the accuracy of machining on metal-cutting machines, in particular during finishing (± 10 microns), are most typical for aviation instrument making and assembly. The processing technology is characterized by small batch size and multi-product variety.
It is noted that due to a rather slow decrease in the correlation function of centered deviations in the dimensions of parts, there is a potential for predicting changes in dimensions for effective control of accuracy by adjustment methods by compensating for the deterministic and correlated components of deviations in the dimensions of the machined parts.
The possibility of using self-learning systems based on measuring and computing equipment in the PMG is substantiated, which would ensure adaptation to the current processing conditions and the choice of the most effective option for a given accuracy and minimum losses of machine time for control. Self-learning is achieved through the use of the target function of the process of control and management, consisting of a linear combination of expressions of accuracy and loss of productivity for control, as the most important characteristics of machining processes that determine the applicability of control and measuring equipment [3].
It is obvious that taking into account only the maximum number of identical signs of deviations for a number of cases of combinations of signs is not always optimal in terms of the criterion of using the information obtained and cannot ensure the achievement of the control goal. In many cases, the second approach to the formation of a variable in magnitude correcting increment seems to be more acceptable, associated with the introduction of a new parameter that takes into account the number of alternations of deviations of neighboring sizes in the combination of signs of a sliding sample, as well as the correlation of the sequence of signs of deviations in the sizes of the controlled process.
Do'stlaringiz bilan baham: | 
